Jiaxiang Ren, Ph.D. Research Proficiency Presentation: 'Optical Coherence Tomography based Imaging and Denoising'

Friday, July 23, 2021 - 2:00pm to 4:00pm
Zoom - contact events@cs.stonybrook.edu for Zoom info.
Event Description: 

Optical coherence tomography (OCT) is a fast and non-invasive optical imaging technology with high spatiotemporal resolution. Its functional imaging approaches have shown promise for both biological studies and clinical diagnoses. For example, Doppler OCT can quantify blood flow velocity without contrast agents and angiographic OCT is increasingly being recognised as the gold standard to diagnose retinal diseases. These advances in OCT permit in vivo visualization which is better for observing the actual experiment effects than in vitro study. However, the imaging process is susceptible to micromotion caused by heartbeat, respiration, and mechanical vibration. Such noisy artifacts result in poor visualization and biased quantification.

In this report, we start with a brief introduction about the OCT system and mathematical principles. The in vivo imaging artifacts are illustrated with two OCT applications. To improve imaging quality, we introduce the denoising and segmentation models in supervised or unsupervised manners. The results in the real world data show that our method can correct most noisy artifacts in various experiment settings.
Computed Event Type: